COMPARATIVE STUDY OF CREDIT CARD FRAUD DETECTION USING HIDDEN MARKOV MODEL

The use of credit cards has drastically increased in today's world because of the increasing advancement in the e-commerce industry. Credit cards are used for offline as well as online purposes and with its uses come the fraudulent activities associated with it.In this paper, we will model the sequence of operations in credit card transaction processing using a hidden Markov model (HMM) and see how it can be used to detect frauds. HMM is initially trained with the behavior of the user or the cardholder. If credit card transaction is not accepted by model with sufficiently high probability, it is considered to be a fraud transaction. It is ensured that the genuine transactions are not rejected. Detailed experimental results are shown to prove effectiveness of our approach and the model is compared with other techniques already available in the real world.Fraudsters are so expert that they come up with new ways to commit fraud each day which demands constant innovation for its detection techniques as well. Many techniques that already exist in the market are based on Artificial Intelligence, Data mining, Fuzzy logic, Machine learning, Sequence Alignment, decision tree, neural network, logistic regression, naïve Bayesian, Bayesian network,Metalearning, Genetic Programming etc